Hell froze over this week. Did you notice? Linus Torvalds gave a keynote address at Spring Comdex, Apple Computer's market share reached 12.9 percent, and Compaq, mighty Compaq, stumbled.

Linus, principal developer and godhead of the Linux operating system, made a appearance that reminded me somewhat of the late Mother Theresa. He called for us to all be our best technoselves. This is easy for Linus to say, standing astride an empire that exists somehow without revenue or deadlines. Linus can call to our inner programmer because he is not facing an unreasonable shipping schedule, an unyielding customer base, or a ruthless sales force. Linus exists in a state of grace within, not of, the software industry. More power to him, though any more power and he'll be walking on water and making blind people see.

So the dealers came to listen (remember Comdex is the Computer Dealers' Exposition), heard Linus call for more love and less competition, then they went out for a smoke, scratching their heads. It was like David Byrne headlining at the Grand Old Opry.

Linus is right, of course. If only we took more pride in our code or, for that matter, even understood what part of the product our code was used for, the world would be a better place. Better certainly than my recent situation of using the Psychic Friends Network for technical support. Attempting a technical breakthrough, I then bought a magic 8-ball to help me make IT decisions. It gave me conflicting answers to every question I asked it, except for one. I asked it several times if I would grow a sixth toe on each foot, and it consistently answered in the positive.

Comdex itself was a yawn. More important news, by far, was being made at the Internet World and NAB shows happening at the same time in different cities. Comdex, at least Spring Comdex, doesn't mean a thing anymore.

If Linus's appearance at Comdex was stirring, the Apple news is even more amazing. I have written many times before that a distinct hardware/software platform without at least 10 percent market share is amazing. And so, I thought, was Apple. But now the iMac is the top-selling personal computer in the world and all the market research companies have Apple busting well past 10 percent. Even better, a significant number of those buyers — around 35 percent in some surveys — are not only new to the Mac platform, they are new to computers altogether. It's the Internet, stupid, but it's also Steve Jobs and a revitalized Apple. I am dumbfounded and amazed. But I also fear Apple has been eating its seed corn (gutting R&D to maintain earnings). Only time will tell. I'm not predicting anything.

Most amazing of all is the resignation or firing of Eckhard Pfeiffer, CEO of Compaq. This is a guy who could do no wrong a year ago, done-in by a $500 million quarterly profit shortfall. There has been lots of speculation about what caused Pfeiffer's fizzle, with some blaming rock-bottom PC prices, others blaming Compaq's delayed decision to market PCs directly a la Dell, and some saying Compaq just doesn't get the Internet. I put the blame elsewhere, on difficulties with the Tandem and Digital acquisitions.

Pfeiffer wanted Compaq to be big, as big as IBM. So now with sales over $40 billion, Compaq is big, and unlike many big companies, the company is also ruthlessly efficient. That's an easy byproduct of growing up in the shadow of IBM. Part of Compaq's fast-growth policy was based on acquisition, especially Tandem and Digital, which were to make the Houston company into solid competitors in big iron — mainframe computers.

These were both prime acquisitions. Tandem Nonstop computers run almost every business you can imagine that can't afford to stop running. NASDAQ, the stock market, literally lives inside a Tandem mainframe. I remember talking with a Compaq executive shortly after the Tandem purchase and hearing him point out with a snort that Tandem used 4,000 salespeople the year before to sell 110 systems. There were obvious opportunities, he thought, for improving efficiency.

Not. Selling mainframes is not like selling PCs. You can't sell a mainframe at CompUSA and you can't sell mainframes at all without massive handholding. Compaq gutted Tandem and Tandem stumbled.

In the case of Digital, the company had stumbled long before being bought by Compaq. What Compaq bought was Digital minus what I think was the company's best asset, its networking division, which had been sold-off some months before. Compaq overpaid for Digital and has been suffering ever since. This will all work out in a few months, I'm sure, so despite Pfeiffer's demise I generally see the recent Compaq stock drop as a buying opportunity. But don't think of Compaq as another high-flier like Dell: Pfeiffer's love of corporate mass and his admiration of IBM have built the company into just another slow-moving corporate giant. That is unlikely to change now, no matter who runs Compaq.

If Compaq paid too much for Digital, we have all been paying way too much for Windows, too, thanks to Microsoft's desire to make beta testers pay to find bugs. Here's the most recent financial analysis courtesy of Paul Braun from Valparaiso, Indiana. Paul should really get out more often.

If you paid $90 for the Windows 95 beta, then paid another $90 for the retail upgrade, and then paid an extra $30 or so for the Plus Pack, which included stuff that should have come with the original release, that's $210 for an obsolete OS. Add another $90 for the beta of Windows 98, which is really Internet Explorer 4 and a bunch of bug fixes for the program you've already paid $210 for. Another $90 scores the retail upgrade of Windows 98, and now you might have to pay another $90 for Windows 98 SR1 to fix all the bugs in the program you've already spent $390 on. Then you are expected to pay another $90 for the beta of Windows 2000 and probably another $90 for the retail upgrade which is really just some extra gizmos pasted on top of Windows 98. The grand total for staying Windows current yet avoiding the really major expense of upgrading to Windows NT is $570, all of which explains why Bill Gates — not Bob Cringely — is the richest man in the world.